Neural Tissue Engineering involves the use of three-dimensional biomaterial scaffolds for transplantation of neural tissue or for induction of regeneration of severed peripheral or central nerves. There is a paucity of information regarding the optimization parameters for the design of 3D constructs from the mechanical, morphological, or charge perspective such that neurite extension is maximized. In addition, cytoskeletal organization and response of neurons in 3D matrices to various stimuli is not well understood. Our laboratory has been using a combination of electron, force, light, video and confocal microscopy to address these issues. Custom surface rendering image reconstruction techniques were used to reconstruct the three-dimensional structure of neurons extending processes in 3D as visualized by confocal microscopy. We have previously reported that low concentration gels of hydroxyethylated agarose are capable of supported neurite extension from a variety of neural cells in vitro.

Agarose hydrogels were fabricated in concentrations ranging from 0.75%-3.0% (w/v). These highly hydrated gels were prepared for electron microscopy using a variety of sample preparation protocols including freeze drying, and critical point drying and their morphological structure and pore radius was computed using scanning electron microscopy.

Near Field Scanning Optical Microscopy (NSOM) is the newest member of the Scanning Probe Microscopy (SPM) family, which includes Scanning Tunneling Microscopy (STM) and Atomic Force Microscopy (AFM). Like its predecessors, it uses a sharp tip and a very precise positioning system to scan the surface of samples. The important difference is that the NSOM tip is made of an optical fiber, with a sub-wavelength size aperture at its end, which is brought to near-field proximity to the sample. In the near-field regime, the resolution is only limited by the size of the light source (the aperture), and not by the wavelength of light, thus allowing NSOM to obtain images with subwavelength resolution. Applying NSOM to biological samples is very appealing, since biological research has developed a wide variety of fluorescent dyes which can be directed with high specificity to objects of interest. Moreover, biological research relies to a large extent on conventional fluorescence microscopy, which is diffraction limited. Application of NSOM to biological problems requires the ability to image rough and soft samples in liquid, a task that presents many challenges.

We have developed a NSOM which is capable of imaging cells in liquid with a resolution of ∼150 nm. This has enabled us to image the organization of fluorescently labeled proteins on the membrane of cells in liquid, and demonstrate for the first time that the proteins are clustered.

Transfection of mammalian cells can be enhanced by forming complexes between exogenous DNA and cationic peptides which facilitate DNA transport through cell membranes. Although complexes of plasmid DNA and cationic peptides have been characterized by microscopy, little is known about the complexes formed between oligonucleotides (ODN) and cationic peptides or how the ultrastructure of these complexes affects the intracellular delivery via the cell membrane. Because ODN have potential as therapeutic reagents in the treatment of many diseases, there is a need to develop improved methods for effective delivery of these molecules. By analyzing molecular interactions between cationic peptides and ODN and their effects on cell membranes, cell viability, and DNA delivery, improved delivery compounds can be designed. In this paper, we compare the interactions of a 28 base ODN with three cationic peptides: a novel compound, MNLEK (H-Met-(Nle-Lys4)7-Nle-NH2), an established transfection reagent, poly(L-lysine) (∼ 54 kDa), and a recently described cationic peptide which delivers both plasmid DNA and antisense ODN to cells, KALA (WEAKLAKALAKALAKHLAKALAKALKACEA).

For high resolution scanning electron microscopy (HRSEM), DU145 human prostate cancer cells were grown overnight on poly-(D-lysine) coated, UV irradiated silicon chips.

Nano-mechanical properties of composite biomaterials are governed by the morphological features and chemical distribution of heterogeneous multiple subunits. However, very little is known about such structure-function correlation at the nanoscale. We have used wool fibers as a model of composite biomaterials for such study. Wool fibers provide an ideal two component cylindrical model for such study as their ultrastructure and chemical composition have been well characterized, they have a core cellular component surrounded by an outer cell layer. The core is 16-40 micron in diameter and rich in axially aligned keratin microfibrils. Outer cells have multiple laminar layers, 60-600 nm thick and are distinctly rich in disulfide bonds. In this study, we have used an atomic force microscope (AFM) to examine the micromechanical properties of the different structural components using complementary techniques of nano-indentation and force-volume imaging. AFM images of transverse sections of wool fibers were obtained in ambient environment (Figure 1).

The exo-cuticle showed a significantly higher mechanical stiffness than the internal cellular core region (Figure 2).

Gel networks that mimic the properties of naturally occurring mammalian elastin have been synthesized for the development of novel biomaterials. Hydrogels with controllable and predictable mechanical properties are pursued for use in drug encapsulation and delivery systems, as scaffolds for tissue engineering and for the development of prosthetic implantable devices. The techniques of genetic engineering and microbial expression have been used to produce a 90KD recombinant protein based starting material that was polymerized into a solvent swollen gel network by the introduction of covalent bifunctional crosslinks at regularly spaced lysyl residues. SEM methods have been used to describe solvent and temperature dependent structural features.

The polypeptide precursor to the elastin-mimetic gels, poly[(VPGVG)4VPGKG], was obtained through bacterial expression (pET system) of an oligomerized gene coding for tandem repeats of the monomer. Gel networks were formed by crosslinking the protein with a bifunctional NHS-ester in both water and anhydrous DMSO. For conventional below-lens SEM studies, samples were dehydrated in an ethanol gradient, critical point dried from CO2(l), mounted on stubs with carbon tape and sputter coated with a 4-6 nm layer of Au/Pd.

Adherent cells generally construct the immediate substrate upon which they reside. This may occur via synthesis and secretion of new materials and/or by rearrangement and modification of existing substrate. The response of adherent cell types to an existing substrate can be influenced by a number of factors which include both the chemical and physical nature of the substrate. Cell adhesion, proliferation, differentiation and death can all be substrate dependent. Much effort has been directed toward chemical modification of substrates to regulate one or more of the parameters noted above. A significant, but somewhat smaller, degree of attention has been paid to the effects of the topography and microtopography on the cell response to substrate materials. Studies to date strongly suggest the topography is a significant factor in cell-substrate interactions. As noted above, it is most probable that both the chemistry and the structure of a substrate simultaneously influence the cellular response. However we wished to determine, particularly for artificial substrates, the role which microtopography can play in cell-substrate interactions.

Every time we take a breath, we are inhaling the results of twentieth century combustion technology. Combustion processes generally produce a multitude of soot and other sub micron sized particulates. The human lungs, via the process of cilia movement expel most of these particles; others are broken down with the aid of macrophage agents. These macrophages absorb particles and incorporate the constitute elements into our bodies. These elements maybe expelled, or they may remain in the body and accumulate over time, as is the case with certain heavy metals. Limited prior research on ‘single-particle’ interaction with lung or bronchial tissue has been conducted. Related research has focused on the statistical significance of soot inhalation on the lung tissue of rodents and primates [1]. Using the methods of single particle examination, founded by previous research into single particle source allocation [2], the examination particles of from human lung and bronchial tissues was performed.

Research on the particle characterization shown here is based on the application of an innovative method developed at MIT, which utilized high resolution transmission electron microscopy (HRTEM), scanning transmission electron microscopy (STEM) coupled with electron energy loss spectroscopy (EELS) and energy dispersive X-ray analysis (EDX).

The efficiency of implantable electrodes for cardiac pacing depends on the ratio of the conductive surface area to the geometric area of the interface with excitable tissue. New models of heart pacers require reduction of post-pulse polarization, i.e. the potential left on the electrode / tissue interface after a pacemaker pulse. Increasing the conductive surface area is an effective method to this end. Microscopy provides an important tool in elucidating the role of surface structure in electrode performance.

Three different surface textures were characterized on a 90% platinum(Pt)/10% iridium (Ir) polished electrode substrate of roughly 5 mm2 geometric surface area. These consisted of the polished substrate itself, a thin film of textured platinum in the 1-3 micron size range, and a sub-micron platinum black coating. Sample surface effects were characterized via scanning electron microscopy (SEM), in-vitro electrical impedance spectrum analysis, and polarization after-potential measurements.

RecA has been the most intensively studied enzyme in homologous genetic recombination. Actin is one of the most abundant and conserved eukaryotic proteins, and is central to muscle contraction, cell motility, and the control of cell shape. Both form helical filaments, and we have been studying both for many years, primarily using negatively stained specimens. We and others have used cryo-EM of frozen-hydrated RecA and F-actin samples, but the resolution obtained by this method is actually worse than that obtained from negatively stained specimens. The criterion for resolution that we have used is the match with an atomic model for F-actin, since this is a much better standard than statistical measures such as Fourier ring correlations or phase residuals, which can only measure the internal consistency of the data.

F-actin is a flexible structure, both with respect to bending and torsional motions of subunits. The RecA filament is 10 times more flexible than F-actin. The bending of RecA and F-actin require that images be straightened, and the ability of a user to fit a spline function to the filament axis limits the precision of the method. The method is only valid for small perturbations from a rigid rod, and thus will introduce artifacts for real filaments. In order to straighten filament images obtained in ice, one needs sufficient contrast (obtained by a large defocus) to be able to apply such algorithms. The torsional motions of the actin subunits, which may be as large as 5-6° per subunit, limit the resolution that can be obtained as one averages over long lengths.

The identification of tubulin as the major protein component of microtubules (1) occurred at about the same time, a little over thirty years ago, as the first demonstration of the ability to obtain 3- dimensional data from electron micrographs. DeRosier and Klug's paper on image reconstruction of the bacteriophage tail (2) marks what many consider the beginning of electron crystallography - determining protein structures by electron microscopy. Microtubules have long been the target of active study by light and electron microscopy, and even before the discovery of tubulin played a significant role in driving technical developments in EM. The demonstration that tubulin could form two-dimensional crystalline sheets (3) seemed to pave the way for high resolution .structural studies. The discovery of the sheet polymer of tubulin came shortly after Unwin and Henderson's first publications on bacteriorhodopsin showing secondary structure resolved by EM (4). Application of their methods to tubulin in the late 1970's resolved low resolution features of the protein, but technical problems prevented achieving much more than showing the molecular envelope. Many developments in both tubulin studies and electron crystallography followed this work over the next 15 years. As techniques were refined for purifying, labeling and visualizing tubulin and microtubules, the dynamic properties of microtubules and their role in the cell cycle were characterized, and some hints about the mechanism of polymerization and disassembly were derived from biophysical and microscopic investigations.

We have reconstructed the three-dimensional quaternary structure of the complete 480 kDa insulin receptor (IR), complexed with NanoGold-labelled insulin, via sets of electron micrographs obtained by low-dose low-temperature dark field scanning transmission electron microscopy (STEM).

Insulin binding to IR in mammalian cell membranes is essential for its manifold effects such as glucose homeostasis, increased protein synthesis, growth, and development. IR belongs to the superfamily of transmembrane receptor tyrosine kinases that include the monomeric epidermal growth factor receptor (EGFR) and platelet-derived growth factor receptor (PDGFR). In contrast, IR and its homologues IGF-1R (insulin-like growth factor 1 receptor) and IRR (insulin receptorrelated receptor) are sub-types of this family that are intrinsic disulfide-linked dimers of two αβ heterodimers. Monomeric receptor TKs are inactive, but are activated by ligand-induced dimerization that results in autophosphorylation. IR-like TKs are also inactive even though they are already dimeric, and are activated by ligand binding without further oligomerization. Insulin binding to the extracellular domain of IR results in autophosphorylation of specific tyrosines to initiate an intracellular signal transduction cascade. However, because the quaternary structure of IR is not known, the structural basis for the mechanism of IR activation by extracellular insulin binding has not been elucidated.

The insulin receptor was purified from human placenta. Bovine insulin was derivatized with NanoGold at the B chain Phel, a location not directly involved in receptor binding. Binding of derivatized insulin to the purified receptor was reduced only slightly compared to binding of the native insulin.

Recent improvements in digital image acquisition, and automated TEM, have lowered the electron dose requirement for collecting a tomographic tilt series to the point where cryo-electron tomography is now feasible. Theoretical calculations indicate that resolution values of 2-5 nm can be obtained for fully hydrated specimens 100-500 nm thick. This resolution is sufficient for identifying individual macromolecules, and macromolecular assemblies, within-organelles and small cells. However, initial applications to cells and organelles fell short of theoretical resolution limits, indicating further development is necessary before cryo-electron tomography reaches its full potential for characterizing macromolecules in situ. Axonemes are ideal test specimens for cryo-tomography because of their appropriate size (diameter of 250 nm), ease of isolation, and unique construction from regular arrays of well-characterized components. The latter include microtubules, radial spokes, central sheath components, outer dynein arm, inner dynein arms, and the microtubule subunit lattice. These components are convenient benchmarks for evaluating the actual detection limit of tomographic reconstructions within a resolution range of 3-30 nm. Furthermore, axoneme components are sufficiently well ordered that computed diffraction patterns contain layer-lines that are useful for judging the quality of individual 2D images.

The past decade has seen major advances in the analytical capability and utility of both fixed beam and scanning beam electron microscopes. In particular, scanning transmission electron microscopy (STEM) and energy-filtering transmission microscopy (EFTEM) have benefited from the development of devices and techniques—including improved electron optics, sensitive solid-state detectors and new software for imaging and electron energy loss spectroscopy (EELS)—that optimize detection of weak spectroscopic signals arising from biological specimens while minimizing specimen damage. Here we discuss and illustrate some of these advances, especially in the context of structural imaging, detection limits and mapping techniques for the biologically important elements phosphorus and calcium. Analytical microscopy of biological tissues is absolutely dependent on cryotechniques. It is generally agreed that rapid freezing and subsequent low-temperature processing, e.g., cryosectioning or direct cryotransfer of frozen-hydrated specimens, is the most reliable way to preserve the native distribution and organization of biological structures. Equally important, however, as an adjunct to spectroscopic analysis is the use of established low-temperature, low-dose techniques for recording optimized images. By limiting beam exposure, low-dose methods greatly improve the quality of images from fragile, freeze-dried preparations. In this case, the quality and information content of, e.g., cryosections are virtually as good as conventional preparations (Fig 1).

Cryo-electron tomography can be used to obtain highly-detailed 3-D reconstructions of micrometer-sized biological structures which are unfixed, unstained and embedded in vitreous ice. Important practical considerations for acquiring the requisite tilt-series images are the choice of electron dose and defocus value, and the effects of contamination and ice thickness.

It is essential to reduce the total dose enough to prevent excessive ice radiolysis and damage to the specimen itself. To reduce the dose, a sensitive cooled CCD camera and microscope automation system are all but required. Because the specimens are thick, it is advantageous to use a high accelerating voltage, and/or energy filtering to reduce inelastic scattering. At high accelerating voltage, it is important to keep the dose high enough so that the signal is above the “scintillator noise” generated by the interaction of the high-energy electrons with the phosphor scintillator of the CCD camera. Initial test reconstructions indicate that scintillator noise, unlike electron shot noise, cannot be mitigated using dose fractionation. Fortunately, we find that large objects embedded in a thick (ca. 1 μm) vitrified ice layer are more resistant to beam damage (as well as charging) compared to smaller objects embedded in ice 100-200nm thick.

Electron microscopic tomography is providing important new insights about the internal structure of the mitochondrion. In particular, the infoldings of the mitochondrial inner membrane (cristae), which are usually rendered as lamelliform baffles, are revealed to have considerable tubular nature. Rather than opening wide to the peripheral compartment (between the inner and outer membranes), the cristae connect to the outside and to each other through narrow (20-30 nm) tubular segments, which can be hundreds of nanometers long. This suggests that diffusion of ions, metabolites and proteins between the intracristal and intermembrane spaces may be restricted.

The earlier tomographic reconstructions were done on conventionally prepared, plastic-embedded specimens, which raises the usual concerns about structural preservation. More recently, we have undertaken tomography of isolated rat-liver mitochondria that have been embedded in vitreous ice (by plunge-freezing in iso-osmotic buffer without chemical fixatives or stains). These frozen hydrated specimens are imaged with a JEOL 4000FX equipped with a Gatan cryo-transfer holder and a Tietz automated data collection system, with a Ik × Ik CCD. For 3D reconstructions, images were recorded at a dose of 5 e−Å2 at 2° increments over the range +/− 60°.

When resins are ultrathin sectioned they are compressed in the sectioning direction. The amount of compression depends very much on the sectioning angle on one hand, and on the properties of the resin on the other hand. Shortening of the sections in the sectioning direction is in the range of 5 to 35%. Even more compression occurs during cryosectioning of vitreous samples at very low temperatures (-170°C); the range was measured to be 30 to 60% (1).

To reduce these compression phenomena, there is basically only one approach possible. The sectioning angle has to be reduced as much as possible. In the past it was shown that by reducing the knife angle from 45° to 35°, a reduction of compression by appoximately one third was achieved (2). The problem with diamond knives is, that the smaller the knife angle, the more susceptible the cutting edge becomes. In practice angles smaller than 35° are not used because of this limitation.

Field emission scanning electron microscopy (FESEM) provides a direct view of a biological sample with high spacial resolution. Combining FESEM with cryo-techniques, macromolecular structures have been obtained from single SEM image successfully (Hermann & Müller, 1992; Chen et al., 1995, 1997). This protocol is now applied to study the extracellular matrix of the diatom A. longipes.

Stalk, which is originated from the terminal region of diatom cells, consists of 3 regions: a surfaceadhered pad, a shaft that separates the cell from the substratum, and a ring-like collar (Fig. 1). Investigating the structure and function of the stalk will help to understand the adhesion and interaction of diatom with the external environment. The extracellular matrix of diatom is highly hydrated, it is difficult to preserve them for EM. It will be easily extracted and collapse during specimen preparation (e.g. fixation, dehydration). In order to preserver the stalk structures close to their native condition and observer them with high fidelity, cryo-preparation techniques are implemented. Cryo-technique is widely used in the plant biology. However, to our knowledge, it has not been used in the diatom research.

The cells of protein-storing plant tissues, including the aleurone of monocots and the cotyledons of dicots, have been widely used to study biological phenomena. The cells of protein-storing tissues share many characteristics including the presence of protein storage vacuoles (PSV) which act as reservoirs for the proteins, lipids and minerals required for germination. Many proteins synthesized in these tissues during germination are transported to the PSV. As a result, protein-storing tissues have become a model system for the study of protein transport.

Research in our laboratory focuses on the protein-storing barley aleurone layer. The aleurone layer is notoriously difficult to prepare for transmission electron microscopy (TEM). Aleurone cells are surrounded by thick cell walls that are difficult to penetrate with fixatives and resins. To increase penetration of chemicals, most of our studies are performed on protoplasts, aleurone cells that have had their walls enzymatically removed. Protoplasts, however, are sensitive to the osmolality of the fixation medium and solutions must be closely monitored during specimen preparation. The development of cells in the aleurone layer adds to the difficulty of preparation for TEM.

A structure preservation of biological and organic samples, close to native state, can only be reached by cryo immobilization techniques. Cryo immobilization allows not only to preserve the high structural integrity but also to arrest dynamic processes in the μs- ms range.

After freezing the sample and preparing the surface of interest, it is important to prevent the sample from ice crystal damage, removal of structural water, condensation of water or other contaminants until imaging. Therefore, ideally the samples are kept below the recrystallisation temperature of water (< 147K) during the transfer from the preparation environment into the imaging chamber.

For the transfer of frozen samples several concepts have been followed in the,past: a) the specimen after manipulation/preparation were submersed in liquid nitrogen and transferred to the cold stage of the microscope or b) a preparation chamber was permanently attached to the microscope column allowing the direct transfers between the preparation chamber and the cold stage in the microscope. These concepts allow either a high grade of flexibility combined with a high risk of contamination or to prevent contamination but combined with inflexibility. In addition the later also does not allow using the microscope during the specimen preparation procedure, nor transferring the specimen to an other imaging device.

It should be no secret by now that ultrarapid freezing is a superior method of specimen preparation for many biological EM projects, and that high pressure freezing is the most versatile of the freezing methods. While cryopreservation is not necessary for all EM studies, it is the method of choice for high resolution work and where “fixation artifacts”, such as distorted membranes, or extraction of the cytosol is a problem. It is true that the machines are expensive, and not always immediately available, but for some questions it is the appropriate technology to use. In the U.S., there are machines available for general use in the far West (Berkeley, CA), the upper Midwest (Madision, WI and Minneapolis, MN), and in the Northeast (Albany, NY). For locations of any of the other 8 machines around the country, interested users can call Technotrade, International at (603) 622-5011.

Current Status of High Pressure Freezing: A decade ago, Studer et al. wrote an article entitled: “High Pressure Freezing Comes of Age” that illustrated how high pressure freezing (HPF) had become a proven technology, useful for preserving ultrastructure with unmatched fidelity in cell types that had previously been difficult to fix well for electron microscopy (EM).